New analysis of Ofcom 2020 data

As society has responded to the challenges of Covid-19, issues of digital inequality have become abundantly clear – be that lack of access to digital tools for home schooling or access to services online. Importantly this is not simply about those who are ‘offline’ and those ‘online’ but highlights that many citizens use digital systems for quite limited purposes. This is due to limited access or having limited digital skills, or both. For example, looking at the situation just before Covid-19 according to the Ofcom Children’s Media Use and Attitudes Survey:

23.4% of 5-15 year olds in the poorest households (NRS Grades D&E) do not have access to both a educationally useable device (laptop, desktop or tablet) and broadband. This equals 524,871 UK children, of which 74,225 are likely studying for GCSEs.

If we then look at the children who can only access a shared device or don’t have access to broadband, the numbers rise quite dramatically. There is therefore a sliding scale of access and use. This circumstance leads to very different capability to respond to Covid-19 situations – in this case home schooling. This pattern holds for all aspects of digital access and use, across all ages.

Much government policy remains focused on material access (availability of broadband) and those who are “offline” (non-users). There is also an assumption that once citizens have obtained access to digital systems and media or digital skills that they will continue to remain “users”.

However, evidence from both the UK and USA (see various Ofcom and Pew surveys) indicates that access can vary over time and over the life course. For example, households might lose access due to ISP or mobile costs and termination of contracts. Longer term, current users may cease to use some or all digital systems at key life stages. This is especially marked in post-retirement, with their digital skills therefore becoming obsolete as technology changes.

These issues make clear that assessing digital inequalities and their consequences requires a deeper understanding of this reality – that digital inequality includes but is not just about being offline.

Types of users

Over the last five years, in collaboration with Good Things Foundation, colleagues and I have been developing an approach to measuring and monitoring these issues.¹ A key part of this has been the identification of citizens who are “limited” users of digital systems. Our analysis is based on the data collected each year by the Ofcom Adults’ Media Literacy Survey. This year’s analysis once again found seven groups, as shown in the chart below.

How did we define users

In our research, we don’t start off with a definition of what is an Extensive user or a Limited user – although of course Non-users are predefined as those people who don’t use digital devices and systems at home or elsewhere.

Our analysis looks at the 17 digital media and systems “Uses” measured by the Ofcom survey. Using a method called “Latent Class Analysis” we group the survey respondents according to their answers. The analysis looks to group the respondents according to the similarity of their responses. It creates groups that have similar “probabilities” for each of the “Uses”. This does not mean everyone in the group is the same – just that the people in the group are likely to be most similar. The graph below shows the result for the six groups of user types (therefore excluding the non-users).

By doing the analysis this way, we avoid imposing a definition of the groups. Instead, we identify the groups and then try to understand them and their characteristics. By doing this we can try to see how groups change over time. This approach also allows us some level of relative measure – these are Limited users compared to the rest of the population. Though in fact, the levels of engagement for Limited users have changed little over time.

What can we say about limited users

We’ve been particularly interested to understand more about the different types of “Limited” users, because these are people who are most likely to be overlooked in the development of policy and practice to address digital inequality.

The analysis identifies three types of “Limited” users in the results who we view as being at risk of differing levels of digital exclusion and inequality. These are:

• “Social and entertainment media only” users (17%). This equates to around 11 million adults. These narrowly focused users are more likely to be younger people (under 35) who have left school at 18 or before, are in lower skilled work and poorer households and live in urban areas.
• “Limited (social media)” users (17%) and “Limited (no social media)” users (10%). This equates to around 17 million adults. These groups are demographically very similar and are more likely to be older (above 55 years old), have left school at 18 or before, have disability or health issues, to be unemployed or retired and financially vulnerable. They are more likely to be in NRS social grades D&E (households on low or very low incomes). These groups significantly lack confidence in their digital skills. Additional analysis also shows the role of place in shaping access and skills. Two things differentiate these two limited user groups:
  • Those who do not use social media are more likely to be rural and also to have disability or health issues that impact their lives. They are more likely to be in areas with rural deprivation such as in Wales, Scotland and Northern Ireland.
  • Those who do use social media are more likely to live in areas of urban deprivation: East Midlands, Yorkshire and Humber, North East England, and in urban areas in Northern Ireland.
• Non-users (15%). This equates to around 10 million adults across the UK. Non-users do not directly engage with digital systems. Perhaps not surprisingly, characteristics of Non-users are very similar to Limited users but they are generally older, and even more likely to have health issues and to be in social housing.

Why is this important

Understanding the different ways in which citizens access and use the Internet – and looking into the detail of digital exclusion – is key to planning interventions and support, and developing policies which address differences across ages and geographies.

Further research and better national data is required to do this properly. For example, our analysis found that ethnicity alone is not a strong predictor of being an Extensive, Limited or Non-user.

However, the minority ethnicity sample sizes are too small in the Ofcom dataset to enable a deeper investigation into how Black, Asian and minority ethnicity intersects with other characteristics – such as low income, geography, age, disability or health issues – in relation to digital inequalities. This is a significant gap in data and understanding.

This analysis makes clear that there is a need to understand the capabilities – skills, equipment and context – of users in order to best help and support them. At a time when the youngest workers, as well as the oldest workers, have been identified as most vulnerable to the economic fall-out of COVID-19, it feels particularly important to look more closely at the interventions needed to support the “Social and entertainment media only” users – where young people with few qualifications are disproportionately represented, so they can get the skills they need for the workplace.

Notes

For the full details of our work please see: Yates, S.J., Carmi, E, Lockley, E., Pawluczuk, A., French, T., Vincent, S., (2020), “Who are the limited users of digital systems and media? An examination of UK evidence”. First Monday, Vol. 25, No. 6. https://firstmonday.org/ojs/index.php/fm/article/view/10847

Yates, S.J., Lockley, E., (2018), “Social media and social class”, American Behavioural Scientist, Vol. 62, Issue 9, pp.1291-1316 https://journals.sagepub.com/doi/10.1177/0002764218773821

Yates, S.J., Kirby, J., Lockley, E., (2015), “Digital media use: differences and inequalities in relation to class and age”, Sociological Research Online, Vol. 20, Issue 4, http://www.socresonline.org.uk/20/4/12.html ↩

User types from previous years

These tables chart the proportions of user types from prior studies

2018

2015

2013